Application of Coded Targes and Video Images in Modeling Body Parts Using Shape From Silhouette

One of the ways to overcome human disabilities is the use of artificial parts, for which modeling plays an important role. This process is usually is done manually and, thus, is time consuming and error prone. Human’s body lacks suitable texture. Thus, the modeling technique should be one that does not require such information. Shape From Silhouette is a technique of such which does not require texture for matching, thus is adopted in this study. The problem with SFS is that it requires many images to form a complete detailed model. This in turn needs several camera positions, the availability of which is both expensive and difficult. To overcome this problem, in this paper it is suggested to use videos as an alternative source. As using video leads to difficulty in orientation of images, coded targets are used to simplify and automate the process. Experiments carried out to test the proposed system, suggest that not only the accuracy of measurements is achieved but also the process is simple and practical enough to be used in practical projects.